470 CHAPTER 8. LABORATORY MEASUREMENT AND ANALYSIS
depends on the accuracy and precision provided by the instrument, coupled with the operator’s dependability and skill (Hudson, et al. 1979, Sharp
1981).
“Precision” refers to the degree of mutual agreement among independent measurements of a single (invariant) physical quantity when measurements are repeated . “Accuracy” refers to the agreement of the measurements with the absolute or true value (Hudson, et al. 1979). For example, if
we make repeated pressure measurements with a bottom-mounted pressure
transducer in a wave tank with motionless water, we can get an indication of the transducer’s precision under these circumstances. However, the
accuracy of the transducer for static measurements is unknown until we
compare measured pressure values with those calculated from theory or
measured with other instruments of known accuracy.
Instrument Calibration
“Calibration” of an instrument is the process of determining the relationship between the instrument’s output signal and the value of the physical
quantity being measured. The calibration procedure usually results in a
mathematical relationship (preferably linear) between the sensor’s output
and the physical quantity. Some instruments require only an initial calibration at the factory, whereas other instruments need to be calibrated
whenever environmental conditions change or at the start of each new experiment. A good quality instrument system should be capable of being
calibrated so its output produces answers that are within 1% of the true
value over the selected working range (Hydraulic Laboratory Techniques
1980).
Instruments that require calibration should be calibrated on a regular
basis, even if they are known to be well behaved (Mansard and Funke 1988).
Although this may appear to be wasted effort, the added comfort of being
secure in your measurements more than offsets the inconvenience.
Measurement Errors
The extent to which measurement errors impact laboratory results depends
on the magnitude of the error relative to other experimental errors. If the
particular experiment suffers from laboratory and scale effects, these error
sources may cause variations that far exceed any instrument error. For
example, known difficulties in scaling sediment transport in movable-bed
models are more likely to skew results than a 3% error in measured beach
profiles.
On the other hand, measurement errors in precise experiments that are
free of adverse laboratory and scale effects must be eliminated or reduced to
depends on the accuracy and precision provided by the instrument, coupled with the operator’s dependability and skill (Hudson, et al. 1979, Sharp
1981).
“Precision” refers to the degree of mutual agreement among independent measurements of a single (invariant) physical quantity when measurements are repeated . “Accuracy” refers to the agreement of the measurements with the absolute or true value (Hudson, et al. 1979). For example, if
we make repeated pressure measurements with a bottom-mounted pressure
transducer in a wave tank with motionless water, we can get an indication of the transducer’s precision under these circumstances. However, the
accuracy of the transducer for static measurements is unknown until we
compare measured pressure values with those calculated from theory or
measured with other instruments of known accuracy.
Instrument Calibration
“Calibration” of an instrument is the process of determining the relationship between the instrument’s output signal and the value of the physical
quantity being measured. The calibration procedure usually results in a
mathematical relationship (preferably linear) between the sensor’s output
and the physical quantity. Some instruments require only an initial calibration at the factory, whereas other instruments need to be calibrated
whenever environmental conditions change or at the start of each new experiment. A good quality instrument system should be capable of being
calibrated so its output produces answers that are within 1% of the true
value over the selected working range (Hydraulic Laboratory Techniques
1980).
Instruments that require calibration should be calibrated on a regular
basis, even if they are known to be well behaved (Mansard and Funke 1988).
Although this may appear to be wasted effort, the added comfort of being
secure in your measurements more than offsets the inconvenience.
Measurement Errors
The extent to which measurement errors impact laboratory results depends
on the magnitude of the error relative to other experimental errors. If the
particular experiment suffers from laboratory and scale effects, these error
sources may cause variations that far exceed any instrument error. For
example, known difficulties in scaling sediment transport in movable-bed
models are more likely to skew results than a 3% error in measured beach
profiles.
On the other hand, measurement errors in precise experiments that are
free of adverse laboratory and scale effects must be eliminated or reduced to